Blog: Policy, Culture, and Resourcing

As part of the RSECon23 Training Event, there was a theme on ‘Policy, Culture, and Resourcing’. This theme started with a talk from Simon Hettrick in the morning where he detailed the difficulties with and pathways toward changing opinion in the academic landscape. Following this in the afternoon we had a discussion session where we generated action points, ideas, and potential problems for how to generate movement.

Our discussion resulted in the following action points:

Evidence

To be able to drive the discussion we will create evidence of the type of training RSEs perform and the benefits it conveys to the research community. Types of evidence could include, REF2021 Submissions, successful Doctoral Training Programs, and Hidden Ref submissions. Additionally, we need to create a catalogue of all the software training that happens, who performs it, and who funds it.

Engagement

To change the way the Universities, Funders, and other employers engage with training we want to become the decision-makers. To become the decision makers we need to start to gradually change the makeup of panels, boards, and steering groups. By sitting in these decision-making groups we can begin to build RSE Training representation. We note that some RSEs will already sit on these panels and engaging with our community representatives can also be a pathway to influence. To work within existing structures we can cite the REF 2028 which will contain an idea of the environment in which research is done. Advocating that showing excellence in software support will fit into the research environment will add positive weight to our argument.

Evaluation

We looked to other pressure groups or frameworks that have generated engagement and movement within our sector. The Athena Swan and the Teaching Excellence Framework (TEF) have both proved to drive change by laying out a framework which then leads to a grading. We propose we create a framework that details the levels of research computing training/support that employers provide to their staff. This will require communication with the RSE Society to provide backing to any grading.

Other points and benefits of software training

When working in these action areas we had some thoughts about other types of discussion we can have that will generate support for comprehensive software training. Centres such as doctoral colleges in many institutions have, in theory, the ability to run centralised courses and access to PhD students. Each training group should look into identifying any existing structures that exist within their institutions. We discussed how software training will lead to the following coupled efficiencies; Firstly, through training, people will develop codes with more ease and produce better-written code that is easier to pass on and maintain this will decrease the time taken overall for research. Following this, well-written codes will also run faster this reduces computational time, computational time and carbon emission from power use are intrinsically linked and we have the potential to reduce the carbon cost of a given piece of research. Finally, by increasing competency, we will increase the number of researchers able to access and use facilities such as national, regional, or local supercomputers. Increasing the use of these facilities will again provide efficiencies as a shared facility is more economical and sustainable than users owning individually powerful computing. Our final area of discussion was around reputation, institutions who invest in their staff will see a reputational gain. Reseachers will be incentivised to work at these places as they will be sure of good continual professional development. Other institutions or employers will also have a higher regard for places that train their staff, and particularly PhD students, as an institution will be known for producing people who are already familiar with these core competencies. Lastly, if we promote a highly skilled software culture across the UK’s research and teaching institutions it will build the country’s status as a digital leader which aligns with the current government strategy.

Conclusion

To drive policy we have determined that we need to be versed in the discussion arguments that make the case for software training as a benefit to UK HE and research institutions. We then need to follow the three action points, Evidence, Engagement, and Evaluation. If we provide sustained effort behind these action points and convene regualry to make sure we create a unified and sustained movement we have a good chance of providing positive policy change.